Title: Research on exchange rate forecast based on MLR-ELM model

Authors: Yi Peng; Kang He; Qing Yu; Yanan Chen

Addresses: School of Economics, Shenzhen Polytechnic, Shenzhen, China ' School of Economics and Management, Changsha University, Changsha, China ' Management College, Guangzhou City University of Technology, Guangzhou, China ' Management College, Guangzhou City University of Technology, Guangzhou, China

Abstract: This paper introduces a new model to predict the exchange rate. The model is a combination model of the multiple linear regression model (MLR) and the extreme learning machine model (ELM). The RMB-USD exchange rate is the object of prediction. Firstly, the sample data are pre-processed and divided into a training set and a test set; then a linear regression equation is created for the training set. The predicted values of the MLR model and other selected independent variables are the input data of ELM, which is determined by the training set. Secondly, the test set data are tested with parameter set obtained from the training set, and the optimal parameters of MLR-ELM model are determined by the performance of the training set and the test set. Finally, the exchange rate is predicted. The simulation results suggest that MLR-ELM model have a better prediction than the multiple linear regression model.

Keywords: exchange rate forecast; multiple linear regression model; MLR; extreme learning machine model; ELM.

DOI: 10.1504/IJMIC.2023.133180

International Journal of Modelling, Identification and Control, 2023 Vol.43 No.3, pp.249 - 257

Received: 04 Jul 2022
Received in revised form: 13 Oct 2022
Accepted: 16 Oct 2022
Published online: 01 Sep 2023 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article